Introduction

Eloquent area glioma surgery is technically demanding. Maximal resection with minimum neurological deficit affects the long-term functional outcome. Several adjuncts such as neuronavigation, image guidance system, intraoperative magnetic resonance imaging (MRI)/computed tomogram (CT) imaging, neurophysiological monitoring, and three-dimensional ultrasonography (3D USG) have been described to improve the extent of safe resection.1 2 3 However, these are expensive tools and may not be available in resource-deficient countries. A two-dimensional USG (2D USG) is easily available and can be considered as an alternative. The major disadvantage of 2D USG is that the tumor–brain interface may not be clear in each plane. Again, the image quality may get affected due to artifacts toward the end of surgery.4 Adding a contrast along the tumor–brain interface increases the chance of picking it up even with 2D USG. In this study, we used water with air bubbles that created multiple vortices along the tumor cavity and these air bubbles within water acted as a contrast picking up the residual lesion. Again insinuating the probe through normal brain parenchyma adjacent to the corticectomy helps in maintaining the image quality. We discussed our experience of 58 cases of supratentorial eloquent area gliomas where 2D USG was used as an adjunct for achieving maximal tumor resection.

Materials and Methods

Fifty-eight patients with primary supratentorial eloquent area gliomas operated between Aug 2018 and Jan 2020 were included in this study where the tumor resection was guided by intraoperative 2DUSG (IOUSG). Patients operated for recurrent or residual lesions were excluded from this study. Patient demographic profile with clinical details was documented. All patients underwent radiological evaluation that included noncontrast CT scan and contrast-enhanced MRI in the preoperative period. Provisional categorization of tumors broadly in to benign or malignant was done on the basis of radiology. However, advanced radiological evaluation such as functional MRI and MR tractography was not performed in these patients as majority required early surgery due to raised intracranial pressure. Other intraoperative adjuncts such as 5-aminolevulenic acid (5-ALA), 3D USG, or intraoperative MRI were not available. All patients were operated under general anesthesia with IOUSG guidance to determine the extent of tumor resection. A curved phased array probe with 6-13 MHz (Sonosite, Micromax, Bothell, Washington, United Sattes) was used in all cases.

Results

The male to female ratio was 1:2 and their age ranged from 18 to 66 years. Radiological (MRI) diagnosis of high-grade glioma (HGG) was made in 38 and low-grade glioma (LGG) in 20 patients (Figs. 1A, 2A–B, 3A–C). The average tumor dimension was 4.8 × 3.6 × 2.2 cm in axial coronal and sagittal plane. There were 32 lesions close to the sensory motor area, 5 were insular glioma, 3 were on the speech area, 3 in the thalamic region, 12 related to the angular gyrus, and 3 on the occipital region close to the visual cortex. At presentation, 3 patients had hemiparesis, 9 had monoparesis, 2 had aphasia, and 2 had parietal lobe features such as apraxia and Gerstmann syndrome. In addition, headache and seizure were present in 49 patients, warranting early surgery.

Close

Fig. 1 (A) Preoperative magnetic resonance imaging showing heterogenous contrast-enhanced lesion in the left frontal operculoinsular region suggestive of a high-grade glioma. (B) Intraoperative ultrasonography showing mixed echogenic lesion with a hyperechoic rim. (C) After partial resection, the tumor cavity and the residual lesion are visualized. Saline with air bubbles filled the entire resection cavity after complete resection of the lesion. Note the hyperechoic rim formed at the periphery of the cavity due to blood clot and air bubbles. (D) Postoperative computed tomography (CT) scan showing resection of the lesion. (E) Immediate post operative CT showing complete resection of the lesion.

Export to PPT

Fig. 1 (A) Preoperative magnetic resonance imaging showing heterogenous contrast-enhanced lesion in the left frontal operculoinsular region suggestive of a high-grade glioma. (B) Intraoperative ultrasonography showing mixed echogenic lesion with a hyperechoic rim. (C) After partial resection, the tumor cavity and the residual lesion are visualized. Saline with air bubbles filled the entire resection cavity after complete resection of the lesion. Note the hyperechoic rim formed at the periphery of the cavity due to blood clot and air bubbles. (D) Postoperative computed tomography (CT) scan showing resection of the lesion. (E) Immediate post operative CT showing complete resection of the lesion.

Close

Fig. 2 (A and B): Preoperative contrast-enhanced magnetic resonance imaging (MRIs) (axial and sagittal) showing left posterior frontal nonenhanced hypointense lesion suggestive of low-grade glioma. (C and D) Postoperative axial and sagittal MRIs showing complete resection of the lesion. (E–H) Intra operative ultrasound images. Tumor appears uniformly hyperechoic with good tumor–brain interface (E). After partial resection residual lesion visualized and margin of lesion well differentiated (F–G). No tumor echogenicity seen beyond the hyperechoic rim suggestive of complete removal of tumor (H).

Export to PPT

Fig. 2 (A and B): Preoperative contrast-enhanced magnetic resonance imaging (MRIs) (axial and sagittal) showing left posterior frontal nonenhanced hypointense lesion suggestive of low-grade glioma. (C and D) Postoperative axial and sagittal MRIs showing complete resection of the lesion. (E–H) Intra operative ultrasound images. Tumor appears uniformly hyperechoic with good tumor–brain interface (E). After partial resection residual lesion visualized and margin of lesion well differentiated (F–G). No tumor echogenicity seen beyond the hyperechoic rim suggestive of complete removal of tumor (H).

Close

Fig. 3 (A–C) Preoperative magnetic resonance imaging (axial, coronal, sagittal) showing hypointense left posterior frontal subcortical lesion. (D) Intraoperative image showing anteriorly paced corticectomy in the operative field. (E) Intraoperative 2D ultrasonography (IOUSG) image showing mixed echogenic lesion with good brain–tumor interface. (F) IOUSG image showing collapse of the resection cavity with ill-defined residual lesion. (G) IOUSG image showing re-expansion of the resection cavity after injection of saline with air bubbles. (H) IOUSG image showing further progression of lesion and the tumor–brain interface is visualized clearly.

Export to PPT

Fig. 3 (A–C) Preoperative magnetic resonance imaging (axial, coronal, sagittal) showing hypointense left posterior frontal subcortical lesion. (D) Intraoperative image showing anteriorly paced corticectomy in the operative field. (E) Intraoperative 2D ultrasonography (IOUSG) image showing mixed echogenic lesion with good brain–tumor interface. (F) IOUSG image showing collapse of the resection cavity with ill-defined residual lesion. (G) IOUSG image showing re-expansion of the resection cavity after injection of saline with air bubbles. (H) IOUSG image showing further progression of lesion and the tumor–brain interface is visualized clearly.

Discussion

Maximal safe resection is desirable in intraparenchymal lesions that may offer good neurological outcome and improved quality of life.1 2 However, complete excision at times is difficult and may not be possible in lesions located at eloquent brain areas. With improvement in technology, various adjuncts such as intraoperative MRI, neuronavigation, fluorescence, and 5-ALA have been added to neurosurgery armamentarium to improve the extent of glioma resection. Recently, 3D USG with facility for contrast injection is gaining popularity and is considered superior because of its real-time imaging. The image resolution is better in 3D USG and tissue differentiation is superior.3 5 6 However, all these modalities are expensive and may not be available in resource-limited countries. Although 5-ALA helps to delineate the residual tumor during surgery, but it is not widely available and expensive. Only a few neurosurgical centers have the facility for intraoperative MRI and 3D USG due to financial constraints. Although these adjuncts facilitate the extent and safety of intraparenchymal tumor removal but in their absence, a simple 2D USG can be used for intraoperative tumor localization. The patient and caregivers must be counseled about all these advanced adjuncts and their availability issues. Proper explained consent should be obtained prior to surgery for standalone use of 2D USG to avoid litigation in future. A 2D USG is easily available yet underutilized instrument for intraoperative use in glioma surgery. This provides real-time imaging of intraparenchymal lesions and can be used as a cheaper but reliable alternative. The use of 2D USG in glioma surgery has been described in the past with good outcome.4 7

The intraoperative USG image quality is mostly user dependent. Proper image acquisition and interpretation are necessary throughout the surgical procedure. The gap between the lesion and the probe has to be minimum to avoid artifacts. The gap between the brain and the USG probe has to be filled with an ideal medium with similar acoustic properties of the brain.8 9 However, sterile gels may be used as a coupling agent to reduce air artifacts. A phase array probe provides a fan-shaped image covering a wide area at the depth and a narrow sector on the surface.8 Therefore, we prefer a wide craniotomy in relation to the tumor, where probe was placed over normal parenchyma adjacent to the corticectomy. Insinuating the USG through normal brain parenchyma may avoid use of additional layer of coupling agent. The concept of keeping the resection cavity vertical and periodically filling it with saline is an established technique to avoid air artifacts.4 However, probe placement over normal brain away from the corticectomy site and simultaneously filling the cavity with saline reduce the possibility of image distortion due to air artifact.

It is of paramount importance to preserve the adjacent white fiber tracts in eloquent area glioma surgery. Preoperative functional MRI or MRI with tractography helps to identify the relationship of these eloquent structures adjacent to the tumor. Intraoperative use of neuro-navigation, fiber tracking, and awake craniotomy with neurophysiological monitoring are the ideal methods for intraoperative identification of displaced white fiber tracts. In conjunction with an IOUSG, maximal safe resection is achievable.3 5 6 2D IOUSG alone is unlikely to yield equivalent results. Nevertheless, it still is an important adjunct in resource-deficient centers. The tumor–brain interface has to be observed carefully throughout the surgery in order to avoid inadvertent injury in cases of eloquent area gliomas. The image quality of the intraoperative USG largely depends on the probe placement. Larger craniotomy is needed for visualizing the entire operative field with a standard 2D USG phase array probe. Sequential swiping of USG probe in all possible four quadrants enables the surgeon to visualize the tumor margin all around. Imaging through multiple planes thereby helps the surgeon to assess the tumor in three dimensions even with a 2D USG.

With progression of surgery, the irregular margins of resection cavity with blood products produce hyperechoic artifacts.8 At times, the tumor cavity may collapse distorting the image quality, and the risk of injury to adjacent eloquent brain thus increases. 3D USG with contrast at this point differentiates the residual lesion owing to its hypervascularity.3 6 Here, we have injected saline with air that re-expanded the tumor cavity. The vortices produced by this air mixture acted as a contrast that aligned along the tumor cavity differentiating the residual lesion from normal brain parenchyma. A smaller size high-frequency probe is also helpful in delineating the tumor margins.10 But these are expensive and may not be available in resource-scarce health-care systems.

Because of hypercellularity, most of brain tumors appear hyperechoic as compared to normal brain parenchyma.11 Tumor margins were better differentiated when appeared uniformly hyperechoic. In our experience, these lesions form a specific subgroup where the ultrasonic image was maintained even toward the end of surgery. In all seven cases in our series where the tumor was uniformly hyperechoic, intraoperative findings matched well with the follow-up MRI. When the tumor was mixed echoic, the extent resection determined by IOUSG may not match accurately with follow-up MRI as seen in nine of our patients. In these cases, the tumor could be well delineated in the beginning. But with the progression of resection as the tumor tissue becomes thinner, the differentiation of brain and tumor echogenicity was not uniform in all four quadrants. In these cases, the tumor–brain interface was more difficult to appreciate at the end of surgery. Again two of them had significant peritumoral brain edema. As the tumor echogenicity was not uniform, it further adds to the dilemma in identifying its edges. Therefore, we believe the tumor echogenicity pattern on IOUSG largely affects the extent of resection.

In three cases where planned subtotal resection was done, the residual lesion appeared larger in the follow-up MRI. At the end of surgery, the residual lesion may not have uniform thickness and we have assessed the residual tumor (all four quadrants) in two dimensions only. This probably results in a false impression of apparent small residue on IOUSG. Intracavitary linear probe USG would have been better to estimate the residual tumor in planned subtotal resection where all corners of tumor cavity can be assessed in multiple planes.

In some series, the extent of tumor resection as determined by IOUSG correlated strongly with postoperative MRI.7 12 However, some of these patients were operated in awake condition or with intraoperative use of ALA for tumor localization and resection.12 In another meta-analysis, where the extent of tumor resection was assessed, the concordance rate between IOUSG and MRI was 82%.13 Yet again intraoperative image guidance system and ALA-induced fluorescence were used in many of the studies analyzed. In another series, the IOUSG correctly identified the histopathological tumor margin in 66 out of 79 cases (83.5%).14 Identification and resection of the residual tumor close to the eloquent region are difficult without the aid of intraoperative image guidance system and awake mapping.

Limitations

A 2D USG is helpful in locating the intraparenchymal brain lesions. Loss of tumor–brain interface with progression of surgery is a major concern with 2D USG. For deep seated lesions located close to white fiber tracts, stand-alone IOUSG may not yield desired tumor resection. Neuronavigation and neuromonitoring if available should be used in addition.

Conclusion

IOUSG with a wide craniotomy is a reasonably good tool to safely resect gliomas located in eloquent areas. The use of saline with air mixture as a contrast aids in resection especially the cystic ones and cases where the tumor walls tend to collapse. The extent of resection is comparable with follow-up MRI, especially in lesions that appeared predominantly hyperechoic or uniformly hyperechoic. However, 2D USG may not give accurate idea of the volume of residual lesion when subtotal excision is planned.